Multiple Platform FirstDesign guidelines for multiple platform games

Emelie Andersson

Emelie AnderssonSpring 2019Master Thesis in Interaction Technology and Design, 30 creditsSupervisor: Håkan GullikssonExtern Supervisor: Linus DilénExaminer: Thomas MejtoftDepartment of Applied Physics and Electronics

Abstract

This thesis investigates the two gaming platforms PC and Consoleand how the interfaces of games on these platforms could be de-signed more e�ciently making it easier to release games on mul-tiple platforms. In other words, how could the Multiple PlatformFirst method look. Little previous work exist on this problem sothis thesis work gathers information from other industries andalso research on user interfaces in games in general. By look-ing at games running on both platforms di�erent best practisesand common solutions were discovered. A study was conductedtesting di�erent in-game components on users. The componentswere selected to test if the users would accept non-traditionalcomponents since users detect when playing on an interface notintended for the platform. This makes the study very complicatedsince the "best" solution might not work if the users does notaccept it for the intended platform. Concepts were designed tocombine the testing of solutions with the users opinion of the so-lutions. The chosen concepts were researched both in literatureand by looking at present implementations in games. To be ableto user test the solutions they were iterated from low �delity pro-totypes on paper to high-�delity prototypes that were playablein Unity. The prototypes were tested on users and data gatheredthrough Think Aloud comments and questionnaire answers. Thisstudy presents a �rst draft of how a multiple platform approachcan be achieved.

Acknowledgements

I would �rst like to thank my external thesis advisor Linus Dilén at Paradox Interac-tive for his constant feedback and will to aid in the work, it was much appreciatedto always have someone to turn to. Also all the "Good job" compliments really liftedmy spirit, so thank you for being here during this whole process.

I would also like to thank my thesis advisor from the university Håkan Gulliksson foralways having a kind word to say during our meetings. It lifted my spirit and madethe report writing much more fun, thank you for that. And also a special thanks forthe very fast e-mail replies when the panic took over.

A big thank you to all the sta� at the Arctic studio for supporting my thesis workby cheering me on and participating in my user study. You are all awesome! Aspecial thank you to my fellow thesis worker at Paradox, Olle, it was so nice to havesomeone in the same situation to talk to and get feedback from. Our thesis therapysessions was much appreciated.

Then there was my peer-review group. What a relief to have you to discuss withduring the hard times of this thesis. Your kind and supporting words along withyour invaluable feedback has made this thesis what is it today.

Finally, I must express my very profound gratitude to my loved ones for providingme with unfailing support and continuous encouragement throughout my years ofstudy and through the process of researching and writing this thesis. This accom-plishment would not have been possible without them. Thank you.

Emelie Andersson

Contents

1 Introduction 1

1.1 Objective 2

1.2 Paradox Interactive 2

1.3 Scope 2

2 Method 3

2.1 Iteration one 4

2.1.1 General Literature study 4

2.1.2 General Field study 4

2.1.3 De�nition of the problem 5

2.1.4 Concept generation 5

2.1.5 Low Fidelity prototypes 5

2.1.6 Low Fidelity test 5

2.2 Iteration two 6

2.2.1 Concept literature and �eld study 6

2.2.2 High Fidelity prototypes 6

2.2.3 High Fidelity test 6

2.3 Iteration three 7

3 User interfaces in games 8

3.1 User interfaces 8

3.2 Game genres 10

3.2.1 Controls in di�erent genres 10

3.2.2 Strategy games 10

3.2.3 Action Games 10

3.2.4 Role-Playing Games 11

3.3 The PC platform 11

3.3.1 User interfaces for the PC platform 12

3.4 The Console platform 13

3.4.1 User interfaces for the Console platform 14

4 Design approaches for multiple platform games 17

4.1 Experience from other industries 17

4.2 Multiple platform games 17

4.2.1 PC �rst 17

4.2.2 Console �rst 19

5 Evaluation and inspection methods 21

5.1 De�nition of heuristics 21

5.2 Heuristic Evaluation of Playability (HEP) 21

5.3 Principles of Game Playability (PLAY) 22

5.4 Game Approachability Principles (GAP) 22

5.5 Game Approachability Issue De�nition (GAID) 22

5.6 Critique against the heuristic evaluation 22

6 Result 24

6.1 General Findings 24

6.2 Summary of previous work 24

6.2.1 Components 25

6.2.2 Complicated situations 25

6.3 Iteration one 25

6.3.1 Concept generation 26

6.3.2 Concept selection 27

6.4 Iteration two 28

6.4.1 Component literature study 28

6.4.2 Hi-� prototypes 29

6.4.3 Test environment 30

6.4.4 Test group 30

6.4.5 Cursor prototype test 30

6.4.6 Menu prototype test 31

6.4.7 Think Aloud comments from the test 34

6.5 Iteration three 35

6.5.1 Heuristic Evaluation of Multiple Platform Issues (HEMPI) 35

6.5.2 Test of HEMPI 36

7 Discussion 37

7.1 Result Discussion 37

7.2 Multiple Platform First 38

7.2.1 Insights in Multiple Platform First 38

7.2.2 Methods in Multiple Platform First 39

8 Conclusion 40

9 Future Work 42

References 42

A Interview Questions 47

A.1 After each cursor prototype 47

A.2 After both cursor prototypes 47

A.3 After all radial menu prototypes 47

A.4 After each menu prototype 48

A.5 After both menu prototypes 48

Glossary

Console A gathering name for devices such as dif-ferent editions of Xbox, PlayStation and Nin-tendo.

D-PAD A button shaped like a cross that allows clicksin four directions. Available on most consolecontrollers.

Hi-� High Fidelity.HUD Heads Up Display.

Lo-� Low Fidelity.

PC Personal Computer.

UI User Interface.UX User Experience.

1(48)

1 Introduction

Gaming is a common leisure activity for people worldwide. It refers to playing elec-tronic games via Personal Computer (PC), Console or through some other medium.It can be performed alone or together with others and is often a source to relax-ation and distraction. The principles of gaming has spread to other industries aswell and games can today be used in, for example, education. Gaming is a nuancedterm that could refer to both casual and hardcore gaming. Overall in 2018 therewas 2,3 billion gamers worldwide [1]. There are so many di�erent ways to game andthere exists so many gamers all over the world. There is a demand for games fordi�erent platforms, in di�erent genres, to play alone or with friends and games withan unique story and environment. A large industry of studios developing games hascontinuously grown with the demand for games. In fact the turnover of the gamingindustry is estimated to be about over 180 billion US dollars in 2021 compared to70 millions dollars in 2012 [2].

There are many unwritten standards a game has to follow, typically an ordinarygame has to be fun or provide some sort of pleasure for the user. Digging deeper,each genre has some similarities that the user expects to �nd when playing a gamein a certain genre. Besides these standards a game also has to follow the standardsfor the intended platform, a game for PC has to work in a similar way to other PCgames.

Today the gamers exist all over the world and they use a variety of platforms. Com-puters, di�erent console and mobile phones are the most common. Since the mar-ket is divided between these platforms increasingly more companies are explor-ing the opportunity to launch their games on several platforms, so called multipleplatform games [3]. New technologies like the SteamLink and the Google Stadiachallenge the very way Console and PC gaming is separated today and is lead-ing the development towards platform independent gaming, which faces the samechallenges as the multiple platform game User Interface (UI) design is dealing withtoday. Launching a game on multiple platforms expands the available market forthe game but since the di�erent platforms present di�erent prerequisites it makesdeveloping multiple platform games a challenge. It is a trade-o� between the rev-enue from the multiple platforms and the cost to develop for several platforms.Today there are some unavoidable di�erences between platforms that will creatework but could the UI design be done in a more e�ective way?

Traditionally a multiple platform game will follow one of two approaches, it willeither design PC �rst meaning it will target the PC audience as its primary target orConsole �rst targeting the console audience primarily. When attempting to design acommon UI or creating them in a smarter way for several platforms it is important tohave the intended platform in mind since the users seem to detect when they playon a UI not intended for the platform. Chunky interactions, readability issues, tactilefeedback issues and navigation problems are just some of the situations whichcould occur and should be able to be avoided in multiple platform designed UI’s.

2(48)

How could the Multiple Platform First approach look and what would it focus on?How can the problems with multiple platform interfaces be detected in a structuredway?

1.1 Objective

The objective of this thesis is to research the design of UI’s for multiple platformsgames. Furthermore to investigate guidelines on how to work around the trade-o�between the cost of creating double UI’s and if only one interface could work formultiple platforms. The thesis aims to answer the following research questions:

• Can traditional evaluation and inspection methods be modi�ed to identifyand evaluate multiple platform issues?

• How a�ected is the playability of games when using traditionally console com-ponents on PC interfaces?

• How a�ected is the playability of games when using traditionally PC compo-nents on console interfaces?

• How could interfaces be designed to mitigate the di�erence between consoleand PC based on the result from this study?

1.2 Paradox Interactive

Paradox Interactive is a game development studio founded in 1999. Paradox isone of the leading publishers of strategy games for PC. The studio has released abunch of successful games in this genre including Crusader Kings, Europa Univer-salis, Hearts of Iron and Cities:Skylines to mention a few. Today the company havemore than 2 million active users each month and studios in Stockholm, Umeå,Malmö, Delfs and Seattle [4]. The development platform is primarily PC, but thecompany has released games on console and mobile platforms as well. Paradoxis interested to see how the multiple platform development process could changefrom how it is done now and if there are some guidelines that could be imple-mented in the design process.

1.3 Scope

This thesis will limit the multiple platform game scope to games for PC and Console.For the consoles traditional hand controllers will be prioritized.

3(48)

2 Method

The overall structure of the method in this thesis was inspired by Design Think-ing 101 by Nielsen Norman Group [5]. This means that the method will follow thestructure of: Empathize, de�ne, ideate, prototype, test and implement. This is aniterative method, in this thesis the phases have been adjusted removing the im-plement phase and deciding to iterate twice over the chosen steps. In �gure 1 thedi�erent phases and their connections is shown as used in this thesis.

Figure 1: Design Thinking 101. [5]

Empathize: The phase empathize aims to create an understanding of the �eld andforemost of the user. The tools needed to create this understanding was a literaturestudy, �eld study of games and study of gaming related forums were the tools wereused.

De�ne: Using the material from the empathize phase, di�erent conclusions or fur-ther de�nition of the actual problem was drawn.

Ideate: In the ideation phase brainstorming was used to generate concepts thatcould be potential solutions

Prototype: The di�erent concepts was prototyped in di�erent stages. The low-�delity prototypes were created with pen and paper and the high-�delity prototypecreated in Unity.

Test: The prototypes were tested in di�erent phases. The low-�delity prototypeswere tested in workshop sessions with colleagues while as the High Fidelity (Hi-�)prototypes were tested with a usability test with potential end users.

4(48)

2.1 Iteration one

To take a �rst stab at the multiple platform �rst method a literature review, a �eldstudy of games and a look over gaming forums was conducted. Which are the keydi�erences between PC and Console UI? Which previous conclusions exist on UIdesign that is or can be applied to the multiple platform issue?

2.1.1 General Literature study

To create an understanding of the �eld both a literature study and an �eld studyof games was conducted. The literature study aimed to �nd what research existedconcerning multiple platform user interface design. The �eld study had as a goalto �nd existing strategies and also to understand the problem better. The literaturestudy took o� from sites such as:

• Google Scholar

• Research Gate

• Diva-Portal

The articles were selected based on their overlap with UI in games or with evalua-tion methods for games. The literature study had a goal to �nd previous researchon multiple platform game interface design and/or research that could be appliedto that �eld. In addition research on evaluation methods for games was included.

To �nd out what the gaming community thought on the subject another litterateurstudy was conducted in parallel. This study aimed to �nd both professional articlesfrom discipline experts as well as thought from the gamers. This study took o� indi�erent forums such as:

• GamaSutra

• Sapphirenation

• GiantBomb

The forums were included if the post or article concerned multiple platform gameissues or opinions.

2.1.2 General Field study

A �eld study was conducted to look at some interesting games from a multipleplatform perspective. This included both active multiple platform games togetherwith PC and Console speci�c games. The games were chosen by insights from theforums and also by collecting knowledge from within Paradox. The study resultedin a list of games to look at as examples of di�erent design ways.

• Europa Universalis IV

• StarCraft 2

5(48)

• Halo 2

• Zelda: Breath of the Wild

• Fortnite

• The Sims 4

• Dishonored 2

• Dark Souls

2.1.3 De�nition of the problem

Based on the previous research and the understanding of the problem it is veryunlikely that there exist any one answer to how multiple platform UI’s should bedesigned generally. A hypothesis was constructed that general guidelines still couldexist on a lower level. When digging deeper and looking at individual parts of theUI it is more likely that conclusions will be able to be drawn.

2.1.4 Concept generation

After deciding on which level to investigate on, a concept generation was conductedwith a brainstorming session which took of from the literature and �eld study. Thebrainstorming session was conducted alone. The layout of the session was inspiredby McCarthy’s article concerning brainstorming alone [6]. Instead of post-its, assuggested in the article, a white board was used to write the ideas on. Insteadof doing two iterations the session went on until one minute had passed withoutany new ideas had been written. After the brainstorming the ideas were groupedtogether and the nonsense ideas thrown away. This method writes solutions on avery basic level with simple sketches or words which is cohesive with Nielsen andNormans suggestions for the ideate phase [5].

2.1.5 Low Fidelity prototypes

The �delity of a prototype refers to how it conveys feeling of the �nal product. Itis about visual design, content and interactivity. When creating a Low Fidelity (Lo-�) prototype only a few visual attributes are presented such as shapes and visualhierarchy. The content is very restricted and only the key elements of content is in-cluded. Even a Lo-� prototype should have some interactivity were a simple simula-tion can be preformed by a human [7]. These prototypes were very simple sketcheswith the di�erent stages sketched to achieve a simple simulation of the function.The layout and content were kept at a minimum level to convey the situation andfunctionality.

2.1.6 Low Fidelity test

The Lo-� prototypes were tested through workshops with one User Experience (UX)designer and one Game Designer from Paradox. The prototypes were presented forthe same amount of time, then discussed and �nally a voting took place selectingthe top ideas. Which ideas were selected as top depended on the situation whichis why the discussion is important to agree on how to evaluate the ideas.

6(48)

2.2 Iteration two

After selecting the two �nal concepts to be prototyped in to hi-� prototypes a newiteration of the method started.

2.2.1 Concept literature and �eld study

The aim of the concept literature study was to �nd which previous research existedin just those �elds and which games standards, if any, existed. The study was donein the same manner as in the previous iteration

2.2.2 High Fidelity prototypes

Taking the chosen concepts further a Hi-� prototype was created. The Hi-� proto-type should be more detailed and closer to the end product. The interactions anddesign should be realistic and real content used [7]. In this case the the focus wason the interactions and less focus on the exact content and design. This meansthat there could be arguments that this prototype is mid-�delity which means it issomewhere between Hi-� and Lo-� but it will be referred to as a Hi-� prototype inthis thesis.

The Hi-� prototypes were created in Unity in an A/B test manner. An A/B test isa method for comparing two versions of a design or function against each other.The A/B test was chosen since it is very good at deciding which version is the mostsuccessful one. On UXPlanet it is recommended even in a pre-launch state whichis more similar to this situation [8]. The layout of the prototypes were decided intoa city with clickable houses since it was appropriate in both situations for both theconcepts.

2.2.3 High Fidelity test

As suggested by Nielsen and Norman the prototype was put in front of real usersto verify it achieved its goal. Based on the research by Nielsen and the graph ofissues discovered versus the number or participants the number of users that gotto participate in the tests of the Hi-� prototypes was set to ten users [9]. To addadditional value to the tests the method of Think Aloud was applied. According toHolzinger this method only needs a minimum of three users [10].

The usability tests was performed with Think Aloud since it doesn’t require as manyusers to give results [10] in order to get some comparable data this will be com-pleted with a questionnaire but there will not be enough users to draw a signi�cantconclusion. Think Aloud means that the user in encouraged to think loud duringthe test and say what they are feeling and experiencing. Using Think Aloud givesaccess to how di�erent people experience the same task which is generally a veryqualitative result. Think aloud can give very varied result which is why triangulationwith other methods is suggested by Charters [11]. The test was a situation basedtest were the user got to preform di�erent tasks which were designed to exposethe participant to the thing it was supposed to test. Each participant got to try bothversions of both prototypes but the order was randomized between participants tominimize any eventual bias between the order of the prototypes. The prototypeswere tested on ten participants at the o�ce of Paradox Interactive. The participant

7(48)

got to �ll out a short questionnaire about experience with gaming before startingthe tests.

An example of a task was:

• Navigate around in the world and test the controls, try selecting at least onehouse from each part in the city

After completing the tasks on one prototype a short questionnaire was �lled out bythe participant. After completing testing both prototypes a comparative question-naire was �lled out.

2.3 Iteration three

Based on the general literature study di�erent evaluation methods for games hadbeen discovered. One was chosen as a predecessor to a new concept. Taking o�in the research on existing methods a new version was developed. It modi�ed andremoved points from the previous version trying to adjust it to identify multipleplatform issues. The evaluation method was hypothetically tested on games indi�erent stages to evaluate how it potentially could work in reality. This test meantgoing over the points in the evaluation with the constructed Unity prototypes as asource.

8(48)

3 User interfaces in games

User interfaces for multiple platform games are complicated and is a�ected by thestandards correlating with the genre and platform the game is intended for. The in-dividual gameplay of each game also a�ects which possibilities exist when wantingto design the UI. This chapter will look into UI for both the PC and Console platform.Which are the most important platform standards to consider? Furthermore someof the bigger genres will be explained with their corresponding standards.

3.1 User interfaces

User interface is according to de�nition the means in which a person controls asoftware application or hardware device [12]. In games the user interface is themeans in which the player experiences the game. In the word of Desi Quintans:

Put simply, the role of a good UI is to provide relevant informationclearly and quickly and to get out of the way once it has done its job.[13]

User interfaces for games is not a pure science with only proved concepts and rulesbut a combination of art and the science by the lessons learned. Even though itis not a science, some people have done work trying to categorize the elementsin the UI to draw conclusions about how they are designed. A categorization ofthe elements in a UI into four categories is suggested by Lorentzon and Fagerholt[14]. UI elements are categorized into the four categories: Diegetic, Spatial, MetaNon-diegetic. Diegetic elements can be seen by the players avatar in the game.Information is provided in a way that tries to make the player think that the avatarhimself is viewing that very same information, is diegetic. Spatial UI elements triesto keep the player immersed in the game as much as possible, even though theinformation they provide is clearly for the player’s eyes only. An example of thisis in tutorials �ne spatial elements tries to fool the player to think the avatar isreceiving the training even though the text and directions are not visible for theavatar. Meta elements are closely related to spatial ones, but with the distinctdi�erent of sitting on a clearly di�erent plane from the game world. Generally metaelements sit on a 2D plane above the game screen and provide player informationon what is going on, for example blood spatters on the screen when the player ishit. The meta elements tries to keep in line with the visuals and aesthetics. Non-diegetic element are there to purely provide usable information, visualization andclear controls for the players. They sit on a 2D plane above everything else likemeta elements but they also usually clearly stand out from the main view. In �gure2 the four di�erent categories for UI elements are presented.

9(48)

Figure 2: UI element categorization

Turunen conducted a study of UI’s in video games with the conclusion that themost important lessons was: Customization options, Genre consistency, Visual ap-pearance, Balance between e�ciency and e�ects [15]. When it comes to looking atcommon �aws in game UI Yordanov also contributed with his post on Sapphire-Nation listing common mistakes. His list and Turunens �ndings overlap on manypoints. Yordanov lists slow to navigate UI and game immersion which is not men-tioned in Turnunes study [16].

When looking into designing a control scheme for a game one has to take thecomfort of the user into account. Dotsenko [17] writes from his experience as adesigner in game control design. Both common PC and common console controlstoday depend on the use of hands. Human hands have limitations which makescontrols using the thumb and index �nger more comfortable than those using thepinkie. With this in mind Dotsenko suggest "Accessibility Tiers" for both standardPC and console controls. These have arranged the di�erent buttons on the consoleand keys on the keyboard into tiers based on how comfortable and accessible thebutton/key is, this is shown in �gure 3.

(a) PC tiers (b) Console tiers

Figure 3: Controller tiers

10(48)

3.2 Game genres

Since there exists genre standards when it comes to the UI of a game it is interestingto look into some of the most common genres. Paradox Interactive is most knownfor the Grand Strategy genre. There could be some things used in other genres thatcould be applied to this genre or vice versa. Some genres are equally representedon each platform and some are favoured on one platform. To follow the genrestandards it is important to make the game accessible to users.

3.2.1 Controls in di�erent genres

On PC it is the keyboard and mouse that is the standard input control. These canbe used in many di�erent setups to allow di�erent interactions. Some of these aremore common in certain genres. Moving a character or the view is very commonand is a key control in almost every game. On a keyboard there are a few commonways to move. One is to use the WASD-keys to move forward,left, backward andright. This can be utilized both for camera and for moving character, in third personand in �rst person view. The arrow-keys can be used in the same way as the WASD-keys. It is also common to use the mouse to move. One example of this is whenthe background moves when the cursor hits the edge of the screen, edge scrolling.One other way is to drag the background by holding down a key and then movingthe background by dragging the mouse. The mouse can also be used to rotate thecamera or a character. Themouse is often used as a cursor allowing for componentslike psychical buttons in the UI and more information heavy menus. When playingconsole games with a controller, the controller is your only interaction method.Common practise is to use the left joystick to move and the right joystick to lookaround. The D-PAD can also be used for movement but is more common to use tonavigate menus. The bumpers are also a common navigation tool in menus.

3.2.2 Strategy games

With gameplay based on traditional strategy board games, strategy games tend togive players a godlike access to the world and its resources. These games requireplayers to use carefully developed strategy and tactics to overcome challenges.More recently, these type of games have moved from turn-based systems to real-time gameplay in response to player feedback. Some subgenres include: 4X, Ar-tillery, Real-time strategy (RTS), Real-time tactics (RTT), Multiplayer online battlearena (MOBA), Turn-based strategy (TBS), Turn-based tactics (TBT), Grand strategywargames [18]. Strategy games are more common on PC, especially in the TBT, TBS,RTS and Grand streategy subgenres. In strategy games on PC it is common to havea lot of information easy accessible and often in many menus. In console menusthe navigation is a little trickier due to the lack of the cursor. The bumpers andsometimes the use of a button could help to make navigation in menus possible.

3.2.3 Action Games

Action games are just that—games where the player is in control of and at thecenter of the action, which is mainly comprised of physical challenges players mustovercome. Most early video games like Donkey Kong and Galaga fall into the actioncategory. Action games have many sub genres including Shooters and Survivals

11(48)

[18].

Shooters let players use weapons to engage in the action, with the goal usuallybeing to take out enemies or opposing players. Shooters are categorized by theplayer perspective: First-person shooters (FPS) are played from themain character’sviewpoint; Call of Duty, Half-Life, and Halo are good examples. With third-personshooters like Fortnite and Splatoon, the action takes place from a viewpoint wherethe player can see the main character, usually from slightly above and behind.Survival action games have really come into their own over the past few years. Thesurvival horror game Resident Evil was one of the earliest (though a linear game),while more modern survival games like Fortnite take place in open-world gameenvironments and give players access to resources to craft tools, weapons, andshelter to survive as long as possible [18]. In FPS it is common practise to use theWASD-keys to move and then use the mouse to rotate and aim, left click to shoot.In console the left stick handles movement, the right joystick handles rotation andaim and A-button or triggers is clicked to shoot.

3.2.4 Role-Playing Games

Probably the second-most popular game genre, role-playing games, or RPGs, mostlyfeature medieval or fantasy settings. This is due mainly to the origin of the genre,which can be traced back to Dungeons and Dragons and other pen and paper role-playing games. Still, hardcore RPGers don’t discount sci-� fantasy-themed RPGslike Mass E�ect, Fallout, and Final Fantasy, which have helped put unique spins onthe genre. RPG also have a bunch of subgenres including: Action RPG, MMORPG,Tactical RPG, Sandbox RPG [18]. RPG’s exist on both console and PC and with both�rst person and third person perspective. The common controls is to move thecharacter with WASD or arrow keys or by moving the character by clicking with themouse on the new location on PC. On console the joysticks usually do this but insome cases the D-pad could be used. On PC the mouse is clicked to select stu�while on console tha A-button does this.

3.3 The PC platform

The PC as a platform means that the user uses a computer to play games. Thismeans that traditionally the user is close to the screen and interacts with the gameusing a keyboard and mouse. The keyboard gives access to many keys that canbe used in di�erent ways in the game and the mouse provides accurate and fastnavigation over the screen. The computer today is used for so many di�erent activ-ities and gaming is one of them. The computer started as an advanced calculatorand has from there evolved to the computers we have today. Overmars describesthe major milestones in computer game development in his text "A Brief History ofComputer Games" [19]. One of the �rst usages of a computer as a gaming platformwas the game Space War that was run on a PDP1 computer. Up until the 90’s thePC gaming was very limited. By the 90’s the computers were getting more pow-erful. PC had better resolution, more memory, hard disks to store game data anda much higher processor speed. Many games were released during this period intime including: Sim City, Grand Theft Auto and Half-Life to mention a few. Over theyears the hardware of the computer became better and allowed for more advanced

12(48)

graphics and more powerful games. The change in hardware, mostly the ability tostore large �les opened up to saving game and creating recurrent games that didn’treally end. As the development of computers went on the market extended beyondPC and today there are other type of computers as well. The most popular non-PCcomputer is the Apple Mac. When it comes to gaming, PC computers are still themost common computer to use.

3.3.1 User interfaces for the PC platform

The PC platform and its controls allows for rather information heavy and menuheavy UI. Since the keyboard has very many keys, hotkeys in a common usage toallow easy access to information in PC UIs. In PC UI it is also common for theuser to be able to "drag and drop" windows, which means that the user can movethe windows, often menus to a location on the screen of their choosing. Hoveringitems with a cursor to get more information, in the form of tool tips is very common.The general layout is often compact since the elements in the UI are usually closeto each other. One example of a menu and information heavy PC UI is EuropaUniversalis IV [20], shown in �gure 4.

Figure 4: Screenshot of Europa Universalis IV UI. [20]

13(48)

Europa Universalis has a map where one clicks to move the people which are sym-bolising armies. In every menu there are a lot of information an clickable stu� toget deeper into the menu. Tool tips arrive on hover. This type of UI uses the mousemobility and ability to hover over and click in menu, the user can move the mousefreely and move to the desired area of the screen easily. It also uses the fact thatthe user is sitting close to the computer. On a TV screen this much and small textwould be unreadable. Europa Universalis IV is an example of a Grand Strategy game.

An example with the usage of hotkeys on PC is StarCraft 2, seen in �gure 5

Figure 5: Screenshot of StarCraft 2 UI. [21]

StarCraft 2 is an example of a real time strategy game (RTS). The menus are actuallypretty deep but with usage of hotkeys and the mouse it is possible to play this veryintense game. Generally FPS games have much more minimalistic UIs and are lessinformation heavy.

To summarize some of the main points on PC UI:

• Keyboard allows for many hotkeys

• Closeness to screen allows for a heavy amount of text

• Mouse cursor makes movement on the screen easy

• Standard for tooltip is on hover location

• Windows can easily be dragged and dropped with mouse

• Buttons are a common element

3.4 The Console platform

The Console platform was in the beginning only intended for gaming and has in itslater days been developed to support other activities as watching Net�ix on yourXbox for example. This is almost the opposite from the computer that was originally

14(48)

developed for other usages before it became a gaming device. The �rst consoles or“video game consoles” appeared in the 70’s. The Magnavox Odyssey was releasedin 1972. Then a rapid development started with new improved consoles releasingcontinuously. Consoles have so far existed in eight generation, with the MagnavoxOdyssey being part of the �rst generation. In the second generation we �nd theAtari 2600 mostly known for the game Pac-Man. In the third generation Nintendo’sNES was released in 1983-1986. In the late 80’s more actors came into the marketand Nintendo’s Super NES got competition from Genesis, both with over 40 millionunits sold. In the �rth generation the Playstation became the most popular con-sole, released in 1994-1995 it sold over 100 million units compared to the Nintendo64 which only sold 32 million units. In the sixth generation the Playstation 2 wasthe bestseller with 155 million units sold, the best selling console in history. In hisgeneration the �rst Xbox was released. The seventh generation started with the re-lease of the Xbox 360 and followed by the release of PlayStation 3 and the NintendoWii. The Nintendo Wii was the best selling console of the seventh generation. Theeight generation which is still ongoing contains consoles like: Wii U, Playstation 4,Xbox One and Nintendo Switch [22].

The controls for the consoles are di�erent depending on which console is beingused. Starting from one analog stick with only one button, to today’s controllers withmany more buttons and options. There is a large variety in the controllers of today,for example the Wii and Wii U have a very di�erent controller from PlayStation andXbox. The common denominator is that all controllers is a combination of buttonsand analog sticks, but some consoles have motion sensors and/or touchscreens.According to Techspot [23] the Xbox 360 controller is the most common used bySteam players and also the most liked among players. A picture of this controllercan be found in �gure 6.

Figure 6: Xbox 360 controller

3.4.1 User interfaces for the Console platform

The console platform and its controls has joysticks which makes rotation move-ment very natural. The console controller has fewer buttons than the keyboardwhich creates some limitations. The D-pad combined with the joysticks gives manynavigation options. In console UI it is common to have very few buttons and very lit-tle information presented at once. This often create more airy interfaces. In menusit is common to let the user step through it with joystick or the D-pad. Hoveringis not as common in console UI as in PC UI so tool tips for example could appearnext to the menu or in a designated area. Due to the stepping in menus more

15(48)

grid layouts are found in console games. Button hints are common to open dif-ferent menus in di�erent views. In console interfaces haptic feedback in the formof vibrations are used to enhance the experience and also to clearly communicatewith the user the e�ect of their actions. The e�ect of actions are also sometimescommunicated with more animations or rather more powerful animations.

An example of a very simplistic in-game UI is Halo 2. It has very few elements onthe screen and they are pretty sparse. There are basically no in-game menus inHalo, to switch weapons you throw away your current for the new one. The Halo 2UI is shown in �gure 7.

Figure 7: Screenshot of Halo 2 in game interface. [24]

There aremoremenu heavy games on console as well. One is the very popular Zeldafranchise. In the game Breath of the Wild menus plays a big role in the gameplay.When you �nd equipment, food or weapons you have to go to the inventory tochoose them. The inventory is also limited which means you may have to go in andthrow something old out to �t the new equipment or food. When changing weaponthere is a smaller in-game menu. The in-game UI and menu UI are shown in �gure8.

16(48)

(a) Zelda small menu

(b) Zelda large menu

Figure 8: Screenshots of Zelda: Breath of the Wild UI [25]

To summarize some of the main points on Console UI:

• Fewer buttons to work with

• Readability more challenging due to screen distance or screen size

• Joysticks making rotating movement very natural

• More haptic feedback

• Button combinations to reach some elements

• More animations

17(48)

4 Design approaches formultiple platformgames

In this chapter the di�erent routes a multiple platform game can take are sortedout as well as the two main approaches when designing games today. Generally agame is either PC �rst or Console �rst.

4.1 Experience from other industries

In a similar way as a game is either developed PC �rst or Console �rst website designhas a similar way, a website can be Mobile �rst or Desktop �rst. Desktop �rst is thetraditional way where a website is designed to work for desktop and then scaleddown to work on a mobile device. Some aspects and features are optimized fordesktop websites and do not scale down very well. In the mobile �rst approach thedesign for this smaller more limited screen is done �rst and then scaled up. Boththese strategies are equally valid but have di�erent advantages. Gonzales lists thepros and cons with the mobile �rst strategy in his article. The main pros includethat content is prioritized and creates a standardized experience over platforms.The downsides is according to Gonzalo mostly the cost of this way of developmentthat it generally requires a larger budget [26].

4.2 Multiple platform games

There are many examples of games that today run on multiple platforms. Some aredeveloped for multiple platform from the start and some games are ported. Thismeans that they were �rst released on one platform and then redesigned to work onthe other platform. Porting a game provides many challenges, not only the designchallenge to make a comfortable UI. The hardware in each console is di�erent andgenerally doesn’t have the same power as the computer. This makes it a risk thatports can have worse performance than a game intended for the platform. Whenlooking into ports it is the UI design issues that are interesting in this context butit is not the only challenge with ports and multiple platform games.

4.2.1 PC �rst

PC �rst is a design methodology in game design targeting the PC users as the pri-mary players of the game. One example of such a game is Fortnite. The menusystem is designed with the direct input as a priority. This does however not meanthat it has to be a bad console UI just because it is PC �rst. Sinclair [27] analyzessome UI’s in his article on Gamasutra. He is hesitant to call Fortnite PC �rst and

18(48)

instead he wants to say it is PC leaning since it is still good and playable on console.He identi�es some solutions to make the UI work on console, one is to augmentthe existing array of buttons peppered around the screen with icon prompts. Onecritique is that the button prompts are spread all over the screen making the playerhave to scan the entire screen. All in all this port works �ne, the UI is shown in �gure9.

(a) PC UI

(b) Console UI

Figure 9: Screenshot of Fortnite menu UI. [28]

19(48)

Fortnite has a well working UI on console even though the design is PC-leaning. Anexample of a PC �rst approach on console that is not working as well is the consoleversion of The Sims 4. This is an example of a console game with an extreme PC�rst approach. The console game is basically a copy of the PC game which createsa bunch of control issues when playing on console. The Sims 4 has a mouse-stylecursor controlled by the left joystick while the right controls the camera. The Sims 4depends on context switching for the user to be able to move around in the buttonsaround the screen. The console UI of the Sims 4, the same as the PC UI is shown in�gure 10. The Sims 4 demonstrates the most of the negative e�ects a PC �rst gamecan have on console.

Figure 10: Screenshot of The Sims 4 UI. [29]

4.2.2 Console �rst

Console �rst targets the console users as the primary players of the game. Allconsole exclusives are examples of this methodology but some ports also usesthis concept as well. Sinclair describes his motivation to why Dishonored 2 is anexample of the console �rst method. The UI looks identical apart from the buttonhints at the bottom of the screen. The main giveaway that this is a console-�rstdesign is in the favored interaction method-speci�c button/key presses on bothplatforms, where a PC �rst game would emphasize direct input with the mouse, thusleaving the player with only the two mouse buttons to worry about. The downsideof this approach is according to Sinclair the UX burden that is put on the PC playerto read over a bunch of key cues instead of encouraging a more natural direct input.The PC UI is shown in �gure 11.

20(48)

Figure 11: Screenshot of Dishonored 2 UI. [30]

Dishonored 2 has a well working UI on PC even though the design is console �rst.An example of a console �rst approach on PC that is not working as well is DarkSouls. This game was ported to PC more or less directly without changing the UI.The largest problem with the PC port though was the resolution and overall quality.

21(48)

5 Evaluation and inspection methods

When looking to evaluate and/or investigate UI’s there are several methods to try.Je�reis compares four di�erent approaches in his article. The heuristic evaluation�nds the highest amount of problems according to the studies conducted by Jef-freis. [31]. Therefore heuristics will be investigated as a inspection method in thisthesis.

5.1 De�nition of heuristics

Heuristics is any approach to problem solving or self-discovery that employs a prac-tical method, not guaranteed to be optimal, perfect, logical, or rational, but insteadsu�cient for reaching an immediate goal. Nielsen [32] developed a list of heuristicsaimed to evaluate software. These heuristics have been widely acknowledged asa useful tool in software development. When looking into video games, they havesome things in common with software in general but is still a di�erent �eld. Thismeans that the heuristics developed by Nielsen is not directly applicable on games.Instead a further development of heuristics, called game heuristics has emerged.

5.2 Heuristic Evaluation of Playability (HEP)

The evaluation method Heuristic Evaluation of Playability (HEP) is introduced in anarticle published by Desurvire et al in 2004 [33]. The article is based on previousresearch on game heuristics by Federo� [34] and heuristics developed by Nielsen[32]. Guidelines developed by Malone [35] and Clanton [36] cover some parts ofthe HEP. Desurvire presents four categories of game heuristics: game play, gamestory, game mechanics and game usability. The study conducted aimed to compilea comprehensive list of the heuristics and verify them since this had not been doneat the time. A prototype consisting of screen shots from a game in development wasused for the HEP. The HEP guidelines were concluded from existing literature andreviewed by experts prior to the test. In order to get a result for how the methodperforms a standard user test was performed on the same prototype. The issuesfound by both methods were compared and analysed. The conclusion was that theHEP proved e�ective in uncovering playability issues. The number of issues foundby the HEP was greater than the ones found by the user study. The user study issueswere more speci�c to the interface and the issues from the HEP more speci�c tothe game. To note is that the prototype was from the beginning of a process andthus the game play and the game mechanics was not fully developed. The articlewritten by Desurvire et al points out the use of HEP as a tool to predict some ofthe issues from a regular user test. The HEP should be considered a complementto regular user studies.

22(48)

5.3 Principles of Game Playability (PLAY)

Game design is becoming a critical area in the �eld of User-Centered design. HEPhas been found useful but only in limited circumstances. Principles of Game Playa-bility (PLAY) [37] is a broad list of Heuristics developed to be of use as a generalizedfoundation that could then be modi�ed for each speci�c game. This set of heuris-tics are developed especially to help developers during the entire design process.The PLAY principles were created by using current and past research on game us-ability Heuristics and design principles as well as the most up to date informationon superior game design from working top-level game designers. Unlike HEP [33],PLAY recognizes game design as an art and a science [37].

5.4 Game Approachability Principles (GAP)

Desurvire and Wiberg present an extension to their previous work on heuristicevaluation including how to make games accessible to inexperienced players. Theheuristics are intended to support usability testing by focusing on issues of accessi-bility within a game when evaluating and then doing usability testing as a checklistto re�ne the design. In addition, and most importantly, the accessibility principlescan be used to design a good tutorial from the onset of game design. In many casesthe lack of accessibility results in the failure of a game resulting occasionally infailure of studio that would otherwise have produced good and successful games.The principles in the GAP list were developed from previous research and basedon current literature in relation to learning. Usability and playability evaluationwas performed using the heuristic evaluation, focusing on how each accessibilityheuristic was supported or violated and then de�ned the issue [38].

5.5 Game Approachability Issue De�nition (GAID)

This article de�nes the key concept of the methods HEP, PLAY, GAP and later GAID[39]. These principles are: Usability, Accessibility, Approachability and Playability.GAID is an attempt to design a �owchart model based on the HEP, PLAY and GAPmodels. The goal of the �owchart is to be able to identify which one of six commonlyencountered problems in game design that a given problem is most likely to beconnected to. GAID is meant to be used as a tool to ease usability studies by takingthe errors found in such studies and then identify the problem type. GAID willhowever not, at present, be able to give the designer any guidance in how to solvethe problem.

5.6 Critique against the heuristic evaluation

The unstructured nature of heuristic evaluation has also led to several criticismsof the technique. For example, Cockton and Woolrych [40] point out that heuristicevaluation does not encourage people to take a comprehensive view of how soft-ware will be used, and that analysts must pick sample tasks and system featuresat random. Cockton and Woolrych also indicate that using heuristic evaluation,

23(48)

and discount methods in general, can cause evaluators to miss problems and toidentify false problems. Cockton and Woolrych conclude that discount methods arebest suited to use in early design iterations rather than as a summative evaluationtechnique.

24(48)

6 Result

Below the results from the study will be presented.

6.1 General Findings

When looking into multiple platform games and UI’s for these kinds of games it wasquickly realised that no general rules applicable for all games on a high level con-sidering the whole UI could be found. Since each game has so many dependenciesfrom being a game in general, from genre standards, platform standard and theparticular games own circumstances it would not work with general UI guidelines.Guidelines that can be used generally needs to be on a deeper level. After thisconclusion a prototyping work began to create a proof of concept for some deeplevel general guidelines for a component level.

A component in general is de�ned as a part of the whole [41]. In this thesis a com-ponent is a part of the UI. To make sense components are organized into categoriesconcerning di�erent parts of the UI. Some of the main categories are:

• Navigation

• Menus

• Heads Up Display (HUD)

• Windows

From these an example of a component from each category is:

• Navigation - Cursor

• Menus - Radial menu

• HUD - Physical Buttons

• Windows - Dockable windows

6.2 Summary of previous work

Based on the games brought up in the theory a summary was done of the situa-tions creating problems when wanting to design a good interface for both platforms.From existing multiple platform games some interesting components were discov-ered that could be used to prove that some conclusions on UI could be drawn ona component level.

25(48)

6.2.1 Components

With the de�nition of components created after the litterateur study di�erent com-ponents were generated with a brainstorming session described in the methodsection. The generated components are presented below.

• Navigation - Tab navigationLeft and right button hints around top menu row on console

• Navigation - CursorCursor on Console

• Menu - Radial menusRadial menus on PC

• Button - HotkeyPress button to access versus mouse click on PC

• Navigation - Menu NavigationButton to reach level in menu console

• Menu - Grid layout in menusMore spacious layout of items, creates di�cult navigation with D-PAD

6.2.2 Complicated situations

Based on the �eld study of games di�erent situations were discovered which eachpresented an interesting problem. These situations are presented below.

1. Information heavy menus (Console)

2. Multiple windows (Console)

3. Too many layers in menu (PC)

4. Too many clicks for an action (PC and Console)

5. Navigation in world and in menu (Console)

6. Precision of 3D selections (Console)

7. Few available buttons (Console)

8. Context switching (PC and Console)

6.3 Iteration one

Based on the general �ndings the �rst iteration of the method the �rst conceptson components to test were created.

26(48)

6.3.1 Concept generation

The result from the concept generation was suggestions of di�erent componentsto investigate. Each component case was connected to one or several problematicsituations. These are the Lo-� prototypes ideas which very simply gives an explana-tion of the interactions and design of each concept. The list of components/casesare shown below.

• Cursor on console (5,6)

• D-Pad to reach HUD (2,4,5,7)

• Radial menu (1,3,4)

• Cursor in menu (1,4,5)

• Dockable windows Console (2,5,7)

After the concepts were generated they were sketched and further explained as avery low �delity prototype.

27(48)

(a) Cursor on console (b) Cursor in Menu on console

(c) Radial menu on PC (d) Dockable windows Console

(e) D-Pad to reach HUD

Figure 12: Concepts sketched

6.3.2 Concept selection

A group with a game designer and one UX-designer got to vote on which casesthey thought would be the most relevant ones to take to a prototype state. Fromthe Lo-� prototypes each case was discussed and valued. Based on this workshopthe prototypes that will be further investigated on is Cursor on console and Radialmenu on PC. These concepts will be taken in to high-�delity prototypes.

28(48)

6.4 Iteration two

To be able to investigate the deeper level approach some components had to beselected and prototyped to be able to test them on users. Based on the resultinglist from the �eld study of games some cases were created, prototyped and testedwith the goal to prove that some general guidelines could be drawn.

6.4.1 Component literature study

When selecting concepts a more speci�c litterature study was conducted looking atwhat had been done previously with these speci�c components both in researchand in already released games.

Radial MenusRadial menus is a common menu type in console games especially in genres withtraditionally heavy menus such as strategy or RPG. The concept originated fromconsole RPGs, introduced by the Mana series, beginning with Secret of Mana (1993).The radial menu, used in conjunction with the analogue stick, allows quick accessto a large number of menu options with a small number of button presses. Both leftand right analogue sticks are used in conjunction with menus in di�erent gamesbut the left stick is more commonly used. Radial menus have often been used asa solution to the problem of porting PC-based genres to consoles, with prominentexamples including Halo Wars, Command and Conquer: Red Alert 3, Tom Clancy’sEndWar, and Stormrise [42].

Samp [43] conducted a study on di�erent radial menus. He also did a comparativestudy between a list menu and a radial menu to measure the time it takes to searchin the di�erent menus. In his study comparing layered radial menus and layeredlist menus with his speci�c layout the radial menu was faster and more accuratethroughout the test. Samp and Decker [44] investigated how to populate layersin radial menus and how this a�ected the search time. Murano [45] conducted acomparative study between a one layered linear and radial menu. The radial, inthis article called pie menu scored higher in how pleasant the felt the appearanceof the menu was, the comfort using the menu and in ease of search.

Free cursorWhen looking into PC games most of them include a cursor controlled by the mousein some way, this is a very standard interaction in a PC game. On console there isno mouse but analogue sticks that could be utilized to control a cursor. Cursor onconsole has been tried in the past which has created some strong reactions fromconsole players of the opinion that cursor belong in PC games and not in Consolegames. On Reddit one user says that the cursor in Call of Duty: Black Ops, It isusually not a great experience.

One game that really put in an e�ort to go away from this truth that cursor is not forconsole is Destiny 2. In his GDC talk, David Candland [46] explains the thought be-hind the cursor in Destiny 2. Candland et al have designed a cursor that is pleasantto use thanks to gravity and back scrolling. The idea behind this is to be able to de-sign more freely and not be constricted to the classical grid layout in every screenfor console games. When having a cursor that can move around easily over thescreen elements can be placed more spacious which is an advantage for consolegames due to the distance between the player and the screen. Even though Des-

29(48)

tiny invested into making the cursor work it was not well received by all. Some arevery strongly into the opinion that cursors being on PC and not Console. Barker[47]writes in his article:

It is time to kill the cursor in console games - Sammy Barker [47]

Cursors have many advantage even on console but is still generally criticized byconsole players.

6.4.2 Hi-� prototypes

Two prototypes were created in Unity, version 3.4.1, each with two test cases. Thecursor prototype tried a traditional PC component on Console. Cursor navigationis the standard way on PC and this prototype tries to take this to console in twodi�erent cases: A free and �xed cursor case. The free cursor is essentially a mousecursor controlled with the left joystick of the controller with edge scrolling to movethe screen. The �xed cursor is a centered cursor which remains centered all thetime. Here the left stick controls the camera instead moving around the centerof the screen. The Menu prototype tries a radial menu, more common in consolegames, on PC and compares it with the traditional list approach. The radial menuwas implemented with three di�erent navigation’s and the user got to try all threeand then select a favourite used in the comparison test with the list. The prototypesare shown in �gure 13.

(a) Cursor

(b) List menu (c) Radial menu

Figure 13: Hi-� prototypes in Unity

30(48)

6.4.3 Test environment

The environment in Unity was built as a city with di�erent houses and lands. Eachhouse and land was clickable. The control used in the cursor prototype was a Xboxcontroller. In the cursor variant of the setting there were several pods, clusters ofhouses and each of the houses turned red when hovered by a cursor and showeda text for a few seconds when clicked to con�rm the click. The setting was identicalfor both the �xed and free cursor test. In the menu prototype there was only onecluster of houses. Now a mouse and a keyboard was used as navigation tools anda menu was opened by right clicking on a house, no hover e�ect was present inthis prototype.

6.4.4 Test group

The test group was between 20-49 years old with a gender distribution of 60 percentmen and 40 percent women. In the group 60 percent dedicated over 10 hours togaming each week. Everyone but one in the group had PC as the primary gamingdevice but more than half played console games occasionally. The group consistedof ten people.

6.4.5 Cursor prototype test

The cursor test gave a unanimous result, everyone who tested the two prototypespreferred the �xed cursor. When looking at the comments during the test andthe scale answers from the questionnaire there are some interesting �ndings herebesides the answer which cursor was best.

In �gure 14 the summary of the participants rating of the experience of the naviga-tion is shown. The scale was from 1 (Pleasant) to 7 (Frustrating).

Figure 14: Experience with free cursor from 1 (Pleasant) to 7 (Frustrating)

There was a strong correlation between the experience with console and strugglewith the free cursor. The more experienced console gamers had less problemswith the navigation. Most of the participants didn’t �nd the navigation hard butfrustrating. When comparing the ratings of the tasks, which were rated from VeryEasy to Very Di�cult there was less di�erence than when comparing rating with thePleasant to Frustrating scale. This was also caught during the test with exponentiallymore sighs and frustrating sounds during the free cursor test. The hardest thingin the free cursor test was to select the moving object, this created a frustrationwith almost all the users. The other thing was the edge scrolling which confused

31(48)

some participants. When selecting still objects in center screen the free cursorwas more appreciated. There were also some comments on the versatile of thecursor since it could work like a mouse on PC, which was the intention. Many ofthe participant mentioned improvements that would make the free cursor better.The most common suggestion was gravity toward the objects. During the test a fewmeasurements were taken. The average time for selecting the moving object were9.97 seconds with an click accuracy of 41.4 percent while the average time to selectall buildings were 35.6 seconds with a click accuracy of 80.3 percent.

In �gure 15 the summary of the participants rating of the experience of the naviga-tion is shown. The scale was from 1 (Pleasant) to 7 (Frustrating).

Figure 15: Experience with �xed cursor from 1 (Pleasant) to 7 (Frustrating).

The �xed cursor had a quicker start-up with all the participants, and the time ittook before they got used to the navigation was considerably shorter. The hardestassignment, to catch the moving object, was still considered hard but the averagerating was lower than the free cursor test. Themoving around was considered easierthan in the free cursor case. Several people commented that they felt they got morefeedback to the movement of the joystick in this prototype. They liked it when thehad to navigate to areas "outside the screen". During the test a few measurementswere taken. The average time for selecting themoving object were 5 seconds with anclick accuracy of 58 percent while the average time to select all buildings were 28.78seconds with a click accuracy of 87.9 percent. When comparing experience betweenthe free and �xed cursor statistically there is a signi�cant di�erence in the opinionsince the p-value is lower than 0.001 [48], it was calculated to 0.0000878. Whencombining this number with the actual opinions it can be interpreted to that the�xed cursor had a signi�cantly better experience than the free cursor.

6.4.6 Menu prototype test

The menu test had two results, which radial navigation was preferred and then if alist or that radial menu was preferred by the participants.

The participants were not in agreement of the navigation within the radial menu.The �xed navigation was considered "consolly" which was disliked by many. Somefelt that the limitations with the �xed navigation made the navigation very clearand easy to understand. One suggestion that came from many participants was tohave a �xed navigation but to be able to drag the mouse over the circle and end upat that part in the menu. A type of shortcut navigation. The motivation for choosingthe free cursor was that it felt very natural on PC to be able to reach the whole

32(48)

screen. Some pointed out that this however could create confusion if you are in amenu and still is able to navigate the rest of the screen. The limited pointer wasequally appreciated as the free pointer, which is showed in �gure 16

Figure 16: Preferred radial menu navigation

33(48)

When the users got to compare the radial with a list most chose the list as thepreferred menu type. When comparing the individual results from the list and theradial test the participants thought both menus were easy to use but when it cameto a selection the list was still preferred. There were many interesting commentsabout what is important in a menu that the radial was more fun and had manymore graphical and design options. The list was considered more e�cient andmany commented that this was also the most common menu type on PC so it feltmore comfortable. The distrubution of the opinion from the test is showed in �gure17

Figure 17: List VS Radial navigation

The participants got to rate the ease of use after each menu. Looking at those ratingof the list menu and the radial menu both of them averaged an easy navigation.The scale of the rating went from 1 (Very easy) to 7 (Very di�cult). The radial menuaveraged a score of 2.1 making it land in the middle of the easy scale. The list menuaveraged a score of 1.3 landing in the very easy spectrum. The ratings for each menuis shown in �gure 18

34(48)

(a) List menu opinions

(b) Radial menu opinions

Figure 18: User ratings ease of navigation in menus

When looking into the statistical analysis of the di�erence in opinion it is not ascertain as the circle diagram above shows. The p-value is calculated to 0.087 whichonly proves certain trend but not a statistical signi�cance [48]. The list got betterratings so it is pre�ered but there is not a statistical di�erence between the opinionof a radial and a list.

6.4.7 Think Aloud comments from the test

As mentioned in the method section Think Aloud was used in this test which meansthat besides the data presented above with times and graphs there are commentsand insights from the participants gathered during the elapse of the test. The mostcommon comment areas will be presented here along with the ones giving thegreatest insights to the method.

• The edge scrolling a�ected my opinion of the free cursor

• The cursor could be controlled with the left stick and the camera with theright stick.

• Being able to take a shortcut in the �xed radial approach would have im-proved my opinion of it

• The radial was much more fun to navigate in

• The list feels easier

One of the participant said the following about the cursor test:

35(48)

Before testing the prototypes I was more negative about it but itfelt surprisingly natural - Participant 5.

6.5 Iteration three

Based on the research done of heuristics in games [33][37][39][38] a modi�cationhas been made in this thesis focusing only on multiple platform UI problems leav-ing the game story, game design and game play outside the evaluation. The purposeis to �nd the situations where the UI gets in the way of the game instead of help-ing it. Since the PLAY method is an extension of the HEP method it will be thefoundation of this smaller multiple platform evaluation. The evaluation is focusedon only identifying where a multiple platform issue with the UI exists. This exten-sion will be called Heuristic Evaluation of Multiple Platform Issues (HEMPI). Somepoints from the Usability and Control section has been extracted. The usability andcontrol contains 9 heuristics: Documentation and tutorial, Status and Score, GameProvides Feedback, Terminology, Burden on Player, Screen Layout, Navigation, Er-ror Prevention and Game Story Immersion. In some of these heuristics there aresubcategories connected to the multiple platform issues these heuristics will beincluded in HEMPI. HEMPI is a suggested modi�cation of PLAY focusing only onidentifying multiple platform issues without �nding general issues with the game.

6.5.1 Heuristic Evaluation of Multiple Platform Issues (HEMPI)

The points are extracted from the previous method of PLAY [37] with some additionswhich are marked with italic text to separate it from the previous work.

1. Status and Score

(a) B1. Game controls are consistent within the game and follow standardconventions for genre and platform.

(b) B2. Status score Indicators are seamless, obvious, available and do notinterfere with game play.

(c) B3. Controls are intuitive, and mapped in a natural way; they are cus-tomizable and default to industry standard settings.

(d) B4. Consistency shortens the learning curve by following the trends setby the gaming industry to meet users’ expectations. If no industry stan-dard exists, perform usability/playability research to ascertain the bestmapping for the majority of intended players.

2. Game Provides Feedback

(a) C2. Provide appropriate audio/visual/visceral feedback (music, sounde�ects, controller vibration).

3. Burden on Player

(a) E1. The game does not put an unnecessary burden on the player by forexample an extensive amount of clicks to preform an action.

(b) E2. Player is given controls that are basic enough to learn quickly, yetexpandable for advanced options for advanced players.

36(48)

4. Screen layout

(a) F1. Screen layout is e�cient, integrated, and visually pleasing, readableon console and accessible on PC.

(b) F2. The player experiences the user interface as consistent (in controller,color, typographic, dialogue and user interface design).

(c) F3. The players experience the user interface/HUD as a part of the game.

5. Navigation

(a) G1. Navigation is consistent, logical and minimalist.

6.5.2 Test of HEMPI

Since HEMPI is drafted from the points in PLAY [37] it shares its problems. It is notreally applicable early in the design process but is more suited to test a �nishedproduct. It could be very applicable when trying to port a game. Looking at theoriginal version but with the other platform in mind, this could give a pointer towhich interfaces will struggle most when changing the platform.

Using HEMPI early didn’t result in any �ndings. The points aren’t written in such away that it is applicable on a prototype and especially not a prototype of a compo-nent as in this case.

37(48)

7 Discussion

In this section the result from the study are interpreted into additions into the Mul-tiple Platform First design guidelines. The results are then discussed and possibleimprovements or other solutions compared with the chosen ones.

7.1 Result Discussion

Based on the results gathered from the component study several conclusions canbe drawn. The test group preferred the list menu over the radial menu. Amongthe di�erent radials there were no concise result in which one was preferred. Eventhough it seemed like a landslide for the list menu both the comments and thestatistical analysis showed that it was not so. In the comments many said it wasmore fun and creative with the radial and the statistical analysis could not provea signi�cant di�erence between the opinion of the list menu and the radial menu.Comparing the di�erent cursors the �xed cursor was preferred over the free cursorin this test group which could be proven by the statistical analysis. These resultsare interesting since they suggest that a radial on PC could be a viable option sincethere is no signi�cant di�erence in the opinion proven in this study. This study alsosuggest that a when looking into cursor on console a �xed cursor is preferred bythe users and is signi�cantly better than the free cursor. This is within the frameof this study with prototypes. Based on the comments from the users cursor onconsole seems like a viable navigation method on console and based on the studya �xed cursor is the better choice of cursor.

One other very interesting thing that was discovered in this study was the verystrong biases which existed in the gaming community. This was �rst discoveredwhen investigating cursor on console, there was a bias that cursor on console wasbad. This was concluded when reading forums about games and also when talkingwith the sta� at Paradox. This bias will be a very important aspect to investigatefurther in order for the multiple platform �rst approach to be successful. It reallygot the mind spinning and asking why this bias is so strong and where it originatesfrom? Since Console games and PC games have been more separated in the pastthis could have created standardization di�erences causing this bias experiencedtoday.

From the user test many comments submerged and the most common ones wasbrought up in the result section. The most commented thing was the scrolling inthe free cursor prototype. There came suggestions that it would have been a bettercomparison if the free cursor didn’t have the scrolling. But in order to conduct acomparison between the cursor both had to have the same ability, in this case tomove over an area larger than one screen size. There were many suggestions to howthis could be solved, maybe the free cursor would have worked better in a menu ora screen where no movement outside the screen was necessary was the opinion of

38(48)

a few of the participants. One very interesting thing is that another participant alsobrought up that a free cursor in a menu would be a bad idea and that it was in thescreen navigation it could have advantages. These con�icting comments show howdi�cult these problems are, even on such a level where di�erent similar things arecompared. Since so many commented on the scrolling and several suggested thatthe camera could have been moved with the right joystick on the controller whileas the left joystick could continue controlling the free cursor. This is an interestingidea which could have potential but it would involve two movements instead ofone which could be confusing. The other thing many commented on was that theyfelt very trapped in the �xed radial navigation where the pointer moved on a linearound a circle, selecting objects on that line. On console it is not uncommon tohave this sort of navigation but when imported to PC there were strong reactions.This could have something to do with that cursor are so common on PC that theplayers are used to being able to move across the entire screen. The suggestionfrom the users to this feeling was that shortcuts should be possible through thecircle. This was an idea during the prototyping stage but was not implementeddue to the scope of this thesis. This could also be something for future tests andimplementations.

This thesis is the �rst attempt of creating the multiple platform approach. If it couldaid in opening peoples mind and creating ideas on which further steps could betaken I am a happy camper. I hope the results from the cursor and menu prototypewill make some game making teams consider using these for both platforms. Themultiple platform industries is growing since it is becoming more common to playon several platforms. Technologies like the SteamLink and the upcoming GoogleStadia makes this behaviour more common. When the conditions changes how theplayer play the games the industry needs to keep up. I think research like the onesampled in this thesis could aid in creating better games in the future just by beingaware of things such as biases and which components work when. The study itselfgave some really interesting insights and data to be used when developing andinvestigating games in the future. Some things could have been done better whichwould have made the study even better. The test group was also very PC heavy,9 out of 10 had PC as the primary gaming platform which could have a�ected theresults.

Since the web development business are experiencing a similar problem with thedesktop versus mobile it could it smart to keep a watchful eye on their progress withthe problem since their solution will probably provide some interesting insights forthe game industry as well.

7.2 Multiple Platform First

Each part of this study has contributed with knowledge to this new approach, Mul-tiple Platform First. Some content has turned into methods, other into libraries andsome into plain insights.

7.2.1 Insights in Multiple Platform First

A few insights has come to light in this study. These insights are thing important toresearch further in order to develop the Multiple Platform First guidelines.

39(48)

Biases

Biases turned out to be important to this study. There exists a bias between Consoleand PC gamers in how the expect their games to look. This became most obviousin the cursor prototype, both when reading forums and when talking to the sta� atParadox. Cursor on console is perceived as a cheap, not very good solution. Thisshowed up in the test since people were surprised they did not hate is as muchas they thought they would. Developing these guidelines further this would beinteresting to investigate further since an understanding of why there is such astrong bias could be a major part of solving the problems this causes.

General vs Deeper level studies

When attempting to create general guidelines this thesis fell short since it was alimited time and only one person doing the investigations. This does however notmean that such conclusions are impossible to �nd. In a larger study it could bepossible and would be of immense use in the development of multiple platformgames. This thesis investigated guidelines on a component level instead. Thisturned out to be more plausible considering the time frame of the thesis.

7.2.2 Methods in Multiple Platform First

Some results from the study has turned into methods that can be used when en-countering Multiple Platform related issues.

HEMPI

HEMPI is today a draft of a method which could be used to detect multiple platformissues in �nished games or maybe a beta at least. Moving on with it would beinteresting to modify it to be useful in a prototype stage of a game as well. Theremight also be other methods that the heuristic approach used in this thesis.

Component Library

When doing the deeper level investigation on components it turned out to be a veryuseful addition to the guidelines to create a library of the components investigatedthat could be extended with more components in the future. This library could workas an aid when designing user interfaces to know which components could workand how the are perceived by users.

40(48)

8 Conclusion

The purpose of this master thesis was to investigate a new approach as an alter-native to the existing ways of designing games. The following problem statementwas made: How could the Multiple Platform First look?. In order to investigate thisthe following research questions were stated:

• Can traditional evaluation and inspection methods be modi�ed to identifyand evaluate multiple platform issues?

• How a�ected is the playability of games when using traditionally console com-ponents on PC interfaces?

• How a�ected is the playability of games when using traditionally PC compo-nents on console interfaces?

• How could interfaces be designed to mitigate the di�erence between consoleand PC based on the result from this study?

The �rst research question was answered by previous research. Several di�erentscientists have investigated heuristics and how it could be applicable for games.Nielsen’s heuristics have been the base of this research which has formulatedpoints focusing more on how games work which is di�erent from other softwaredevelopment. These methods are themselves rather new and still have not solvedall issues. When adapting the latest version of this research to address only themultiple platform the issues from the previous version follows. This issue is thesame as with Nielsen’s heuristics. It is hard to apply early in a design process onun�nished things, they need to be very Hi-� or implemented in order for heuristicsto come up with interesting �ndings. When testing HEMPI on the prototypes in thisstudy it was quickly realized that they were to low-level to get any results. HEMPIturned out to be a better method for example when a game is to be ported. Goingthrough the points of a PC game but imagining a console controller would certainlybring most of the problems into the light. This however needs to be tested in orderto say for sure but it shows great potential.

Moving on with the other questions the �rst problem encountered was the di�-culty of �nding general conclusions. When starting the litterateur and �eld studythe intention was to �nd these general solutions and suggestions. It was quicklydiscovered that this would not be possible within the scope of this thesis why adeeper level approach was chosen instead. This does however not mean that gen-eral conclusions are impossible to �nd but it will require more time and a very ex-tensive study in order to draw such conclusions. The study conducted went downto a deeper level and investigated di�erent components and how the could beused in multiple platform interfaces. In this thesis only two components were in-vestigated but if a library were created with more components a multiple platformsinterfaces design process could take o� from the component studies and it could

41(48)

aid in the selection of components. When deciding between two approaches in-volving the study’s components, the results can either help steer the decision, oroutright settle it.

In summary all research questions was answered even if some of them were an-swered by asking more or new questions.

42(48)

9 Future Work

There are so much that could be done within this �eld. The study conducted withinthis thesis was limited both in time, scope and participants. Looking into the generalguidelines which this thesis failed to achieve could be very interesting for gamedevelopment in the future. This would probably need much time and resourcessince investigating existing games thoroughly and conducting a large user studywould be viable ways of getting the data for this type of conclusions. Extending thecomponent library beyond the two components explored in this thesis could createa very useful tool when designing multiple platform games and maybe create abigger variation in PC and Console games as well. If a radial is accepted by the users,why not use it when it �ts our game? Taking the heuristic evaluation modi�cationHEMPI further could turn into a helpful tool, maybe even in a ideation or prototypephase. A further investigation of how biases a�ect a gamers appeal to a certaincomponent or UI could be very interesting for making these guidelines even moreuseful.

43(48)

References

[1] Liza Brown. Digital video game trends and stats for 2019. https://filmora.wondershare.com/infographic/video-game-trends-and-stats.html, 2019,accessed 2019-02-28.

[2] Tom Wijman. Mobile revenues account for more than 50 per-cent of the global games market as it reaches 137.9 billion in 2018.https://newzoo.com/insights/articles/global-games-market-reaches-137-9-billion-in-2018-mobile-games-take-half/, accessed 2019-04-23.

[3] Per Christensson. Multiplatform de�nition. https://techterms.com/definition/multiplatform, 2006, accessed 2019-02-28.

[4] Paradox Interactive. Annual report 2017. https://www.paradoxinteractive.com/en/paradox-interactive-ab-publ-publishes-annual-report-for-2017/, 2018, accessed 2019-01-28.

[5] Sarah Gibbons. Design thinking 101. https://www.nngroup.com/articles/design-thinking/, 2016, accessed 2019-04-08.

[6] Brian McCarthy. How to brainstorm alone - try this proven method thatonly takes 10 minutes. https://www.linkedin.com/pulse/how-brainstorm-alone-try-proven-method-only-takes-10-brian-mccarthy/, 2015.

[7] Per Christensson. Prototyping 101: The di�erence between low-�delity andhigh-�delity prototypes and when to use each. https://theblog.adobe.com/prototyping-difference-low-fidelity-high-fidelity-prototypes-use/,2006, accessed 2019-02-28.

[8] CanvasFlip. An ultimate guide to a/b testing on prototypes. https://uxplanet.org/an-ultimate-guide-to-a-b-testing-on-pre-live-apps-4bd57679e8cc, 2016, accessed 2019-04-24.

[9] Jacob Nielsen. Why you only need to test with 5 users. https://www.nngroup.com/articles/why-you-only-need-to-test-with-5-users/, 2000,accessed 2019-03-03.

[10] Andreas Holzinger. Usability engineering methods for software developers.COMMUNICATIONS OF THE ACM, January 2005/Vol. 48, page 71-74, 2005.

[11] Elizabeth Charters. The use of think-aloud methods in qualitative research anintroduction to think-aloud methods. COMMUNICATIONS OF THE ACM, January2005/Vol. 48, page 68-72, 2003.

[12] P Christensson. User interface de�nition. https://techterms.com, 2009, ac-cessed 2019-03-03.

44(48)

[13] Desi Quintans. Game ui by example: A crash course in the good andthe bad. https://gamedevelopment.tutsplus.com/tutorials/game-ui-by-example-a-crash-course-in-the-good-and-the-bad--gamedev-3943, 2013,accessed 2019-03-05.

[14] Magnus Lorentzon Erik Fagerholt. Beyond the hud user interfaces for increasedplayer immersion in fps games. Technical report, Chalmers University of Tech-nology Department of Computer Science and Engineering, 2009.

[15] Jyrki Turunen. The good, the bad and the unpleasant – a study of graphical userinterfaces in video games. Technical report, Tampere University of TechnologyDepartment of Information Technology, 2017.

[16] Alexander Yordanov. What are the best tv and gaming console ux/ui designpatterns? https://www.quora.com/What-are-the-best-TV-and-Gaming-Console-UX-UI-Design-Patterns, 2015, accessed 2019-03-03.

[17] Andrew Dotsenko. Designing game controls. http://www.gamasutra.com/blogs/AndrewDotsenko/20170329/294676/Designing_Game_Controls.php,2017, accessed 2019-03-03.

[18] Vince. The many di�erent types of video games and their subgenres. https://www.idtech.com/blog/different-types-of-video-game-genres, 2018, ac-cessed 2019-03-05.

[19] Mark Overmars. A brief history of computer games. https://www.stichtingspel.org/sites/default/files/history_of_games.pdf, 2012, ac-cessed 2019-03-03.

[20] 2013 Paradox Interactive. Europa Universalis IV. Sweden.

[21] 2010 Blizzard Entertainment. StarCraft II: Wings of Liberty. U.S.

[22] Michael Poe. Evolution of home video game consoles: 1967 –2011. https://www.hongkiat.com/blog/evolution-of-home-video-game-consoles-1967-2011/, 2019, accessed 2019-06-10.

[23] Shawn Knight. Playstation and xbox controllers most popular among steamgamers. https://www.techspot.com/news/76631-playstation-xbox-controllers-most-popular-among-steam-gamers.html, 2018, accessed2019-03-03.

[24] 2004 Microsoft Game Studios. Halo 2. U.S.

[25] 2017 Nintendo. The Legend of Zelda: Breath of the Wild. Japan.

[26] Frederic Gonzalo. Understanding the di�erence between mobile-�rst, adap-tive and responsive design. https://fredericgonzalo.com/en/2017/03/01/understanding-the-difference-between-mobile-first-adaptive-and-responsive-design/, 2017, accessed 2019-04-30.

[27] David Sinclair. Secretly console �rst: A better approach to multi-platformgame ui design. https://www.gamasutra.com/view/news/328311/Secretly_console_first_A_better_approach_to_multiplatform_game_UI_design.php, 2018, accessed 2019-03-20.

45(48)

[28] 2017 Epic Games. Fortnite. U.S.

[29] 2017 Electronic Arts. The Sims 4, Console edition. U.S.

[30] 2016 Bethesda Softworks. Dishonored 2. U.S.

[31] Cathleen Wharton Kathy M. Uyeda Robin Je�ries, James R. Miller. User interfaceevaluation in the real world:a comparison of four techniques. Proceedings ofCHI’91, New Orleans, page 119-124, 1991.

[32] Jacob Nielsen. Usability inspection methods. CHI ’94 Conference Companionon Human Factors in Computing Systemspage, page 413-414, 1994.

[33] Jozsef A. Toth Heather Desurvire, Martin Caplan. Using heuristic to evaluatethe playability in games. CHI 2004 Conference on Human Factors in ComputingSystems, page 1509-1512, 2004.

[34] Melissa A Federo�. HEURISTICS AND USABILITY GUIDELINES FOR THE CREATIONAND EVALUATION OF FUN IN VIDEO GAMES. Technical report, Department ofTelecommunications of Indiana University, Indiana, USA, 2002.

[35] Thomas W. Malone. Heuristics for designing enjoyable user interfaces: Lessonsfrom computer games. CHI ’82 Proceedings of the 1982 Conference on HumanFactors in Computing Systems, page 1-12, 1982.

[36] Chuck Clanton. An interpreted demonstration of computer game design. CHI’98 CHI 98 Conference Summary on Human Factors in Computing Systems, page1-2, 1998.

[37] Charlotte Wiberg Heather Desurvire. Game usability heuristics (play) for evalu-ating and designing better games: The next iteration. Online Communities andSocial Computing, Third International Conference, OCSC 2009, page 557-566,2009.

[38] Charlotte Wiberg Heather Desurvire. User experience design for inexperiencedgamers: Gap – game approachability principles. Game User Experience Evalu-ation, page 169-186, 2015.

[39] Jonas Molin. Gaid: A practical model of game approachability testing of com-puter games. Technical report, Department of Informatics at Umeå University,Umeå, Sweden, 2010.

[40] A Wollrych G Cockton. Understanding inspection methods: Lessons from anassessment of heuristic evaluation. Blandford A., Vanderdonckt J., Gray P. (eds)People and Computers XV—Interaction without Frontiers, page 171-191, 2001.

[41] Cambridge Dictionary. Meaning of component in english. https://dictionary.cambridge.org/dictionary/english/component, 2019, accessed2019-04-29.

[42] Marino. Radial menu. https://www.giantbomb.com/radial-menu/3015-737/,2018, accessed 2019-03-03.

[43] Krystian Samp. The design and evaluation of graphical radial menus. Technicalreport, Digital Enterprise Research Institute, 2011.

46(48)

[44] Stefan Decker Krystian Samp. Visual search in radial menus. INTERACT 2011:Human-Computer Interaction – INTERACT page 248-255, 2011.

[45] Iram N. Khan Pietro Murano. Pie menus or linear menus, which is better?Journal of Emerging Trends in Computing and Information Sciences Vol. 6 No.9September, page 476-481, 2015.

[46] David Candland. Destiny’s tenacious design and interface. https://www.youtube.com/watch?v=zp4NZ8i80QIandfeature=youtu.be, 2016, ac-cessed 2019-03-03.

[47] Sammy Barker. Soapbox: Can we please kill cursors in console games.http://www.pushsquare.com/news/2019/02/soapbox_can_we_please_kill_cursors_in_console_games, 2019, accessed 2019-03-03.

[48] MiniTab. What can you say when your p-value is greater than 0.05?https://blog.minitab.com/blog/understanding-statistics/what-can-you-say-when-your-p-value-is-greater-than-005, 2019, accessed 2019-04-29.

47(48)

A Interview Questions

Below are the questions asked to the participants after completing di�erent stagesof the user tests. The rating questions was answered using Google forms while asthe other questions were answered orally.

A.1 After each cursor prototype

• Please rate how the following worked for you: (1 Very Easy- 7 Very Di�cult)

– Selecting a house– Selecting the capsule– Moving around in the world

• How did you experience the navigation? (1 Pleasant - 7 Frustrating)

A.2 After both cursor prototypes

• Which cursor did you prefer?

• Why did you prefer that cursor?

• How would you rate the overall cursor on console experience (1 Pleasant - 7Frustrating)

A.3 After all radial menu prototypes

• Please rate the ease of navigation from the di�erent radial menus: (1 VeryEasy- 7 Very Di�cult)

– Free pointer– Limited pointer– Fixed pointer

• Which radial navigation was best?

• Which radial navigation was worst?

• Why was the "best" radial best?

• Why was the "worst" radial worst?

48(48)

A.4 After each menu prototype

• Please rate how the following worked for you: (1 Very Easy- 7 Very Di�cult)

– Opening a menu– Selecting an object in the menu

• How did you experience the navigation within the menu? (1 Pleasant - 7 Frus-trating)

A.5 After both menu prototypes

• Which menu type did you prefer?

• Why did you prefer that menu?

• Could radial menu be a viable menu type on PC?